Detection of Single-Axis Pitch Bearing Defect in a Wind Turbine Using Electrical Signature Analysis

Abstract

Recently, multiple wind turbine failure databases have reviewed that pitch bearings are one of the subassemblies with the highest failure rates, and largest contributors to overall downtime. If a monitoring technology is developed and can give an early warning about the pitch bearing condition, the maintenance process can be largely improved; downtime and losses can be minimized. Electrical signature analysis (ESA) has been extensively investigated for years to monitor wind turbine main drivetrain structural health condition, including generator bearings, multi-stage gearbox gears and bearings. ESA has advantages of being low cost, hardware free (no additional sensor is required), has minimum impacts on system normal operation, and can be directly applied to commercial drive units to provide an online remote condition-based monitoring solution. In this paper, ESA is used to monitor and trend deterioration of pitch bearing health condition. Various fault indicators (FIs) has been investigated, a detailed comparison indicates negative-sequence FI is the most sensitive to detect single-axis pitch bearing failure. This hardware-free solution is validated by both simulation and field data from MW-scale wind farms and turns out to be the first field-validated effort to use ESA to monitor pitch bearing health condition, and provide single-axis pitch bearing defect detection, as reported in the literature.